An intelligence report circulated across the US military this spring during the war with Iran claimed that a Chinese ship in the Middle East was carrying components linked to a nuclear weapons program, prompting preparations for a US operation to intercept the vessel, according to CNN.
Four sources familiar with the episode said the military prepared to act. Two said armed US personnel were preparing to board the ship, while military aircraft were already airborne, according to two other sources.
Officials only discovered shortly before the planned operation that the report, produced by an analyst at a special operations command, had been generated with the assistance of artificial intelligence. A chatbot used by the analyst had incorrectly identified the ship’s cargo. CNN could not determine what the material actually was.
One source described the report as “entirely false” and said it “almost started a war.” Any US operation against a Chinese vessel could have escalated into an armed confrontation between Washington and Beijing.
The analyst had queried a chatbot about intelligence concerning the ship’s manifest originating with US Special Operations Command Pacific in Hawaii. The system combined open-source information with classified signals intelligence and reached the erroneous conclusion. The analyst then used AI to package the findings into a standard intelligence report that was disseminated to military officials. It was unclear whether the chatbot was commercial or government-developed.
“The internal tools are mostly just copies of the commercial stuff wearing lipstick,” said a former senior US official familiar with military and intelligence AI systems.
The incident highlights risks as the US military rapidly expands AI use for intelligence analysis, targeting, logistics, budgeting and battlefield decision-making. Defense officials argue AI can accelerate decisions and that Washington cannot afford to fall behind potential adversaries such as China.
In January, Defense Secretary Pete Hegseth unveiled an “Artificial Intelligence Acceleration Strategy,” saying: “We will unleash experimentation, eliminate bureaucratic barriers, focus our investments and demonstrate the execution approach needed to ensure we lead in military AI.”
However, officials said AI adoption remains decentralized, with different systems, orders and safety standards across government and no uniform method for verifying AI-generated information. One source warned that AI targeting is expanding without clear guidance on how human oversight will prevent civilian casualties or fratricide.
Another source said similar AI “hallucinations” have occurred elsewhere in the intelligence community, while some officials fear pressure to produce intelligence faster is increasing the risk of mistakes.
“Young analysts,” sources said, are particularly likely to trust the technology uncritically.
“AI allows you to get to a bad idea faster,” one source said.
By Tamilla Hasanova
